Measuring social class: is a multi-dimensional approach needed?

A good place to start this would be through posing a question:

If an upper-class individual (defined economically) loses their job, do they then cease to be upper-class and become working-class or a member of the precariat (social class below working-class and characterised by precarious employment relations/unemployment)?

Your answer is likely to be no, and this gets to the central argument of this blog post: an individuals’ social class is not defined only through their occupational status, or the status of their parent(s). It is instead a combination of the economic, cultural and social practices and resources they are able to draw upon (Reay, 1998;Savage et al, 2014). Despite the increasing recognition within the social sciences of the multi-dimensional nature of social class there is still little innovation in terms of conceptualising social class in this way in quantitative research aiming to measure social class. This is mainly due to the dominance of the NS-SEC (National Statistics Socio-economic Classification) in national survey and statistical research, which defines social class occupationally, and the absence of any attempts to broaden the conceptualisation of social class in quantitative forms of research.

In this blog, I will first outline the NS-SEC approach to social class and the limits of this approach for understanding social class. Thereafter, I will discuss how a broader notion of social class has been attempted in social research (e.g. Savage et al, 2013) and suggest how this can be taken forward.

The limits of the NS-SEC

The NS-SEC is a way of classifying individuals into social class groups by the occupation they are in. The different occupational catagories, based on the employment-relationship and nature of work, are: (1) higher managerial and professional occupations, (2) lower managerial and professional occupations, (3) intermediate occupations, (4) small employers and own account workers, (5) lower supervisory and technical occupations, (6) semi-routine occupations, (7) routine occupations. The aim of this definition is to treat class as a discrete variable whereby the isolated effect of it can be tested. So that, for instance, we can see how social class background will influence health or educational outcomes (Savage, 2015).

Now, viewing social class like this is limited in many ways. As previously stressed, it fails to consider that what it means to be any given social class is as much social and cultural as it is economic. So, to be upper-class is to engage in certain social environments and with others from similar social networks, as well as having dispositions which promote engagement in certain cultural activities (e.g. playing tennis rather than playing football, for instance). This is not to disregard the role of the economic dimension of social class, but it is to treat it as one of many dimensions which formulate an individuals’ class identity (Savage et al, 2014). Furthermore, such a rigid catagorisation ignores the fluidity of class identity (Dorling, 2014) and implies that similarities cannot be shared across class positions. Lastly, treating social class as a discrete variable also means that its intersectional relationship with concepts, such as gender and ethnicity, are ignored which is troubling during a period where intersectionality is becoming more relevant.

The Great British Class Survey (GBCS)

The first example of adopting a multi-dimensional approach to social class in Britain was in the GBCS (Savage et al, 2013). In this, social class was measured quantitatively through three Bourdieusian class concepts: social, cultural, and economic capital (Bourdieu, 1986). Underpinning the survey was an assumption that “it is possible to draw fine-grained distinctions between people with different stocks of each of the three capitals, to provide a much more complex model of social class” (Savage et al, 2013, p. 223). Social class was measured as:

  • Social capital: measured through respondents’ contact with individuals from 37 occupations listed in the Cambridge Social Interaction and Stratification (CAMSIS) scale.
  • Cultural capital: measured through engagement in two forms of cultural capital – (1) ‘highbrow’ cultural capital and (2) ‘emerging’ cultural capital. The former included respondents’ engagement with classical music, attending stately homes, museums, arts galleries, jazz, theatre, and French restaurants. The latter included the level of engagement with video games, social network sites, the internet, playing sport, watching sport, spending time with friends, going to the gym, going to gigs and preferences for rap and rock. Scores were then created for engagement in each form of cultural capital.
  • Economic capital: measured via household income, household savings and house price. The latter two were combined to create a single ‘assets’ variable.

Now this is still limited, but for different reasons to the NS-SEC conceptualisation of social class (see Sociology). Particularly, this definition of social class failed to adopt a nuanced understanding of the concepts of social, cultural and economic capital. For example, an individuals’ occupation did not form any part of their economic capital, and the differentiation between ‘highbrow’ and ‘emerging’ cultural seems outdated (Bradley, 2014). Moreover, the measure of social capital was a simplistic investigation of ‘who you know’, without any exploration of utility of networks and type of relationships.

However, whilst there is valid criticism to be made, the promise of such an approach to measuring social class should not be ignored. It is perhaps the first attempt at conceptualising social class in survey research in a multi-dimensional way and this should be celebrated, despite the future work required. What is of importance in this future work is that sociologists make their conceptualisations of social class evident and seek to account for the complexity of social class as a relational concept, rather than reverting to simplistic and normative dualisms between good/bad cultural, social and economic practices.

Feel free to comment below and if you are interested in debates around social class then many of the links provided are good starting points, especially Savage et al (2013), Dorling (2014), Savage (2015) and Reay (1998).

Author: Ross Goldstone

Ross Goldstone is an ESRC-funded PhD student at Cardiff University, School of Social Sciences. His research focuses on the relationship between social class and further education (FE) participation and experience.